here's why I look at yards

[Millhouse]

We're better than 47% of teams on defense.
We're better than 25% of teams on offense.

.47*.5 + .25*.5 = 36%. That's what you can expect to win. Works out to a record of 4.3 wins to 7.7 losses. Pretty close.

From a yardage ranking perspective, 65% of those 4.3 wins are attributable to the defense, 35% to the offense. ( 47/(47+25) = 65% ) ( 25/(47+25) = 35%
There's a reason why people look at the offense more this year. It's because being ranked 64th in yards is much better than being ranked 90th in yards.

This is a good predictor. A drawback is that it assumes linearity between rankings - it works less well at the extremes. LSU's and Alabama's defenses are so great, the difference between a 1st and 20th ranked defense is way bigger than a 40th vs 60th ranked defense.

Here's every team's projected 12 game record using that same approach. It's not going to be perfect for every team but it's pretty good. LSU would appear to be a little lucky this year. Their record is better than the projection because their d backs scores so much on d and special teams.

It seems like some people on the board want us to be the opposite of baylor. let's get positive yard differential by expecting top 10 defenses and an offense that just has to be not terrible. i don't think that's realistic and sustainable.

Ok St is pretty far down the list too.

This would work really well to predict if there were more than 12-13 games.

Name Proj W Proj L
Wisconsin 10.9 1.1
Boise State 10.7 1.4
Alabama 10.5 1.6
Stanford 10.2 1.8
Georgia 10.0 2.0
Arkansas State 9.8 2.3
West Virginia 9.8 2.3
Southern Mississippi 9.8 2.3
Virginia Tech 9.5 2.6
Michigan 9.4 2.6
UCF 9.4 2.7
BYU 9.2 2.8
Nevada 9.2 2.9
Georgia Tech 8.9 3.1
TCU 8.9 3.2
Houston 8.8 3.2
Michigan State 8.8 3.3
Oregon 8.8 3.3
Oklahoma 8.8 3.3
Texas 8.6 3.5
Utah State 8.6 3.5
California 8.5 3.5
Notre Dame 8.5 3.5
South Florida 8.5 3.5
Ohio 8.5 3.6
Missouri 8.5 3.6
Texas A&M 8.4 3.6
USC 8.3 3.8
LSU 8.2 3.9
Arkansas 8.2 3.9
South Carolina 8.1 3.9
Virginia 8.1 4.0
Temple 8.1 4.0
Louisiana-Monroe 8.0 4.0
Florida State 7.8 4.3
Clemson 7.7 4.3
Toledo 7.7 4.3
SMU 7.7 4.4
Air Force 7.6 4.4
Army 7.5 4.5
Illinois 7.4 4.7
Northern Illinois 7.3 4.8
Nebraska 7.3 4.8
Cincinnati 7.0 5.0
San Diego State 6.9 5.1
Penn State 6.9 5.2
North Carolina 6.9 5.2
East Carolina 6.7 5.3
Louisiana Tech 6.6 5.5
Oklahoma State 6.6 5.5
Florida 6.5 5.5
Hawai'i 6.5 5.5
Northwestern 6.4 5.6
Tulsa 6.4 5.6
Rutgers 6.4 5.7
Miami (Florida) 6.3 5.8
Vanderbilt 6.2 5.8
Washington State 6.2 5.8
Arizona State 6.2 5.8
Florida International 6.2 5.9
Western Michigan 6.1 5.9
Baylor 6.1 5.9
Pittsburgh 5.9 6.1
Miami (Ohio) 5.8 6.3
Arizona 5.8 6.3
Louisville 5.7 6.3
Louisiana-Lafayette 5.7 6.4
Texas Tech 5.7 6.4
Mississippi State 5.6 6.4
Eastern Michigan 5.6 6.5
Ohio State 5.5 6.6
Tennessee 5.4 6.6
Navy 5.4 6.7
North Carolina State 5.3 6.8
Iowa 5.1 6.9
Kent State 5.1 7.0
Fresno State 5.1 7.0
Utah 5.0 7.1
Western Kentucky 4.9 7.1
Buffalo 4.9 7.2
Bowling Green 4.8 7.2
Purdue 4.8 7.3
Middle Tennessee 4.6 7.4
Washington 4.6 7.4
Wake Forest 4.5 7.5
UCLA 4.4 7.7
Syracuse 4.3 7.7
Iowa State 4.3 7.7
Central Michigan 4.2 7.8
Connecticut 4.1 7.9
New Mexico State 4.1 7.9
Wyoming 4.0 8.0
Oregon State 4.0 8.1
San Jose State 3.8 8.2
Kansas State 3.6 8.5
UTEP 3.5 8.5
Duke 3.4 8.6
Colorado State 3.4 8.7
Kentucky 3.3 8.8
Maryland 3.2 8.8
Marshall 3.0 9.0
Troy 3.0 9.0
Tulane 2.9 9.1
Boston College 2.9 9.2
Auburn 2.8 9.2
Minnesota 2.7 9.3
Ball State 2.6 9.4
Florida Atlantic 2.5 9.5
Indiana 2.3 9.7
North Texas 2.3 9.7
Colorado 2.3 9.7
UAB 2.2 9.9
Mississippi 1.9 10.1
Rice 1.9 10.1
Akron 1.6 10.5
Idaho 1.4 10.6
UNLV 0.9 11.1
Kansas 0.7 11.3
New Mexico 0.5 11.6
Memphis 0.4 11.7

 

[OttoinGrotto]

It seems like some people on the board want us to be the opposite of baylor. let's get positive yard differential by expecting top 10 defenses and an offense that just has to be not terrible. i don't think that's realistic and sustainable.

Great post. Sure looks like you put a lot of time into that.

 

[Millhouse]

Great post. Sure looks like you put a lot of time into that.

not too much. cfbstats to excel vlookups nothign to it

 

[OttoinGrotto]

not too much. cfbstats to excel vlookups nothign to it

Ok then. Either way, I feel the message is important.

 

[Timonen]

i love the point but wonder if there is a way to control for blowouts where yardage stats get screwy. take LSU vs WVU. LSU was killing them so they 1) let WVU pass and rack up yardage and 2) stopped running an offense that would have a chance at compiling yards in favor of one that burned clock.

if we're using yardage as a primary predictor, this probably needs to be worked into the formula.

 

[rrlbees]

Except we didn't play every D1 team. We played 11 of them, which are the ones that made up our D1 record of 4-7

We're better than 27% of the teams on defense.
We're better than 36% of the teams on offense.

 

[Millhouse]

Except we didn't play every D1 team. We played 11 of them, which are the ones that made up our D1 record of 4-7

We're better than 27% of the teams on defense.
We're better than 36% of the teams on offense.

that works out to 3.5 w, 7.5 losses. (really 4 and 7)

so maybe this is how each side comes away thinking o or d is the problem. our defense was better than the offense relative to all offenses and defenses but not the 11 temas we played.

i screwed up in the orig post including the 1AA game. if we just look at d1, 47th percentile D and 25th percentile O works out to 3.96 wins (vs 4 in reality)

final wins and losses works out well either way but you're right about blame - depends on whether you look at all of D1 or just opponents.

 

[ImperialOrange]

Like the numbers, but they dont predict as much as confirm what has already happened. Not sure how you'd use it to predict future outcomes.

Anyway you can work special teams yardage into those equations as well?

 

[Millhouse]

Like the numbers, but they dont predict as much as confirm what has already happened. Not sure how you'd use it to predict future outcomes.

Anyway you can work special teams yardage into those equations as well?

I think they're the best way to predict. I harped on it the last two years before the skids. i'm nothing if not good at harping

 

[007]

Given the title of the thread milly, my first thought was...

Seriously though, good post.

 

[ImperialOrange]

I think they're the best way to predict. I harped on it the last two years before the skids. i'm nothing if not good at harping

So what's your prediction for next season then? Or are those the numbers you posted.

 

[Millhouse]

So what'd your prediction for next season then? Or are those the numbers you posted.

i'll give you a prediction after a handful of games.. give me a few bcs games to work with.

every 20 spots you move up in ranking on either side or combined gets you one more win

this year it predicted 4.3 W - 7.7 L. we were 5-7. (thanks wake and toledo.)

if we want to get to 7-5, we need to move up 54 spots if i did the math right (54 in o or d, 27 in both, whatever combo)

i don't think a nassib offense is going to move up much. it is what it is. no one that dynamic coming in.

so we would need the defense to get back close to 2010 ranking. Unlikely.

our record has been better than what this stuff would predict. downside of that is we might be fooled into thinking we're closer to being good than we actually are.

I say start Hunt and hope for the best. worst case is he takes his lumps in a year that would've been rough anyway.

 

[Dick_in_MI]

so maybe this is how each side comes away thinking o or d is the problem.

I don't take a side on the issue although despite the fact that our offensive woes got the overwhelming bulk of the emphasis, I have maintained all year that our D sucked as well. Neither was THE single reason for our overall malaise. Both were a problem, as can be seen by your stats and BEE's comments.

Your stats do, however, fail to take into account another elephant in the SU room this past season...special teams. They were also a cause of our problems and thus our subsequent record.

Offense, defense, specials...nary a sliver lining in sight

 

[Millhouse]

I don't take a side on the issue although despite the fact that our offensive woes got the overwhelming bulk of the emphasis, I have maintained all year that our D sucked as well. Neither was THE single reason for our overall malaise. Both were a problem, as can be seen by your stats and BEE's comments.

Your stats do, however, fail to take into account another elephant in the SU room this past season...special teams. They were also a cause of our problems and thus our subsequent record.

Offense, defense, specials...nary a sliver lining in sight

we out performed what the yardage would predict even with special teams excluded. if special teams costs us games, we got lucky in other ways to exceed what the formula predicts

special teams is more a fat pig than an elephant in the room

 

[ImperialOrange]

every 20 spots you move up in ranking on either side or combined gets you one more win

this year it predicted 4.3 W - 7.7 L. we were 5-7. (thanks wake and toledo.)

if we want to get to 7-5, we need to move up 54 spots if i did the math right (54 in o or d, 27 in both, whatever combo)

i don't think a nassib offense is going to move up much. it is what it is. no one that dynamic coming in.

so we would need the defense to get back close to 2010 ranking. Unlikely.

our record has been better than what this stuff would predict. downside of that is we might be fooled into thinking we're closer to being good than we actually are.

Good numbers to work with certainly. We could be better on both sides in terms of consistancy and on field play and still not jump in the rankings much due to difficulty of the schedule next year.

We have 2 coaching vacancies so hopefully 1 goes to someone that can coach special teams. ST's lost as many games as any one unit this year. Almost lost us the Toledo game as well.